I can't imagine how the beta could be responsible for this. I would look at the guider before reverting SGPro... star trails are only caused by one thing... improper movement of the mount. Since SGPro never touches the mount during exposure, I'm not sure what part of SGPro could be responsible. It might be one thing if images immediately after some SGPro mount movement are always trailed, but if almost all of them are, it is a long shot that SGPro is to blame. Furthermore, going back to SGPro 2.5.1 and getting better images doesn't prove much of anything because environmental conditions are different (for guiding, etc).
Make sure that you have a proper settling time for PHD2. Some mounts get under the limit and then fly by "0 error" to the opposite side (backlash) and then you have SGPro taking an image while PHD2 is still issuing heavy corrections to get you where you need to be.